html-to-text - npm Package Compare versions

Comparing version 0.0.3 to 0.0.4

example/html-to-text.js

@@ -14,3 +14,3 @@ var path = require('path');
 htmlToText.fromFile(path.join(__dirname, 'test.html'), {
-	tables: ['invoice', 'address']
+	tables: ['#invoice', '.address']
 }, function(err, text) {
@@ -17,0 +17,0 @@ if (err) return console.error(err);

lib/html-to-text.js

@@ -8,3 +8,11 @@ var fs = require('fs');
 
+var helper = require('./helper');
+
 function htmlToText(html, options) {
+	options = options || {};
+	_.defaults(options, {
+		wordwrap: 80,
+		tables: []
+	});
+
 	var handler = new htmlparser.DefaultHandler(function (error, dom) {
@@ -16,4 +24,4 @@
 	});
-	var parser = new htmlparser.Parser(handler);
-	parser.parseComplete(html);
+	new htmlparser.Parser(handler).parseComplete(html);
+	
 	var result = buildText(filterBody(handler.dom), options);
@@ -43,6 +51,2 @@ return _s.strip(result);
 
-function convertHtmlSpecials(text) {
-	return text.replace(/ | /, ' ').replace('€', '€');
-}
-
 function wordwrap(text, max) {
@@ -74,4 +78,4 @@ if (text.length > max) {
 	text = _s.strip(text);
-	text = convertHtmlSpecials(text);
-	text = wordwrap(text, options.wordwrap || 80);
+	text = helper.decodeHTMLEntities(text);
+	text = wordwrap(text, options.wordwrap);
 	return text;
@@ -96,2 +100,6 @@ }
 
+function formatHorizontalLine(elem, fn, options) {
+	return '\n' + _s.repeat('-', options.wordwrap) + '\n';
+}
+
 function tableToString(table) {
@@ -160,5 +168,19 @@ // Determine space width per column
 
+function containsTable(attr, tables) {
+	if (tables === true) return true;
+
+	function removePrefix(key) {
+		return key.substr(1);
+	}
+	function checkPrefix(prefix) {
+		return function(key) {
+			return _s.startsWith(key, prefix);
+		};
+	}
+	var classes = _(tables).chain().filter(checkPrefix('.')).map(removePrefix).value();
+	var ids = _(tables).chain().filter(checkPrefix('#')).map(removePrefix).value();
+	return attr && (_.include(classes, attr.class) || _.include(ids, attr.id));
+}
+
 function buildText(dom, options) {
-	options = options || {};
-	var tables = options.tables || [];
 	function walk(dom) {
@@ -185,4 +207,7 @@ var result = '';
 							break;
+						case 'hr':
+							result += formatHorizontalLine(elem, walk, options);
+							break;
 						case 'table':
-							if (elem.attribs && elem.attribs.class && _.include(tables, elem.attribs.class)) {
+							if (containsTable(elem.attribs, options.tables)) {
 								result += formatTable(elem, walk);
@@ -189,0 +214,0 @@ break;

package.json

 {
   "name": "html-to-text",
-  "version": "0.0.3",
-  "description": "Simple html to text converter",
+  "version": "0.0.4",
+  "description": "Simple html to plain text converter",
   "main": "index.js",
@@ -14,9 +14,11 @@ "scripts": {
   	"underscore": "1.x.x",
-  	"underscore.string": "2.x.x"
+  	"underscore.string": "2.x.x",
+    "optimist": "0.x.x" 
   },
   "keywords": [
-  	"html",
+    "html",
   	"node",
   	"text",
   	"mail",
+    "plain",
   	"converter"
@@ -26,3 +28,6 @@ ],
   	"node": "*" 
+  },
+  "bin": {
+    "html-to-text": "./bin/cli.js"
   }
 }

README.md

@@ -26,3 +26,3 @@ # node-html-to-text
 htmlToText.fromFile(path.join(__dirname, 'test.html'), {
-	tables: ['invoice', 'address']
+	tables: ['#invoice', '.address']
 }, function(err, text) {
@@ -49,6 +49,23 @@ if (err) return console.error(err);
 
- * `tables` allows to select certain tables by the `class` attribute from the HTML document. This is necessary because the majority of HTML E-Mails uses a table based layout. So you have to define which tables should be treaded as `table`. All other tables are ignored.
- * `wordwrap` defines after how many chars a line break should follow in `p` elements.
- 
+ * `tables` allows to select certain tables by the `class` or `id` attribute from the HTML document. This is necessary because the majority of HTML E-Mails uses a table based layout. Prefix your table selectors with an `.` for the `class` and with a `#` for the `id` attribute. All other tables are ignored. You can assign `true` to this attribute to select all tables. Default: `[]`
+ * `wordwrap` defines after how many chars a line break should follow in `p` elements. Default: `80`
 
+## Command Line Interface
+
+It is possible to use html-to-text as command line interface. This allows an easy validation of your generated text and the integration in other systems that does not run on node.js.
+
+`html-to-text` uses `stdin` and `stdout` for data in and output. So you can use `html-to-html` the following way:
+
+```
+cat examples/test.html | html-to-text > test.txt
+```
+
+There also all options available as described above. You can use them like this:
+
+```
+cat examples/test.html | html-to-text --tables=#invoice,.address --wordwrap=100 > test.txt
+```
+
+The `tables` option has to be declared as comma separated list without whitespaces.
+
 ## Example
@@ -81,3 +98,3 @@
 				<td>
-					<table>
+					<table id="invoice">
 						<tr>
@@ -97,3 +114,3 @@ <th>Article</th>
 							</td>
-							<td align="right" valign="top">6,99€</td>
+							<td align="right" valign="top">6,99&euro;</td>
 							<td align="right" valign="top">7%</td>
@@ -113,3 +130,3 @@ <td align="right" valign="top">1</td>
 							<td>&nbsp;</td>
-							<td>to pay: 10,24€</td>
+							<td colspan="3">to pay: 10,24€</td>
 						</tr>
@@ -129,3 +146,3 @@ <tr>
 					<hr />
-					<table>
+					<table class="address">
 						<tr>
@@ -197,8 +214,10 @@ <th align="left">Invoice Address</th>
 ARTICLE                  PRICE   TAXES             AMOUNT   TOTAL   
-Product 1                6,99€   7%                1   6,99€   
-Contains: 1x Product 1                                   
-Shipment costs           3,25€   7%                1   3,25€   
-                                 to pay: 10,24€    
-                                 Taxes 7%: 0,72€         
+Product 1                6,99€   7%                1        6,99€   
+Contains: 1x Product 1                                              
+Shipment costs           3,25€   7%                1        3,25€   
+                                 to pay: 10,24€                     
+                                 Taxes 7%: 0,72€                    
 
+--------------------------------------------------------------------------------
+
 INVOICE ADDRESS          SHIPMENT ADDRESS         
@@ -210,2 +229,4 @@ Mr.                      Mr.
 
+--------------------------------------------------------------------------------
+
 LAW OF REVOCATION
@@ -218,2 +239,4 @@ At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd  no sea
 
+--------------------------------------------------------------------------------
+
 TERMS OF CONDITION
@@ -220,0 +243,0 @@ At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd  no sea

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

27390

increased by29.21%

Number of package files

9

increased by28.57%

Lines of code

290

increased by39.42%

Number of lines in readme file

276

increased by9.09%

Worsened metrics

Dependency count

4

increased by33.33%

Dependency changes

• + Addedoptimist@0.x.x

• + Addedminimist@0.0.10(transitive)

• + Addedoptimist@0.6.1(transitive)

• + Addedwordwrap@0.0.3(transitive)